Arquitectura cognitiva basada en el gradiente sensorial y su aplicación a la robótica móvil

El objetivo de esta tesis es el desarrollo de un sistema completo de navegación, aprendizaje y planificación para un robot móvil. Dentro de los innumerables problemas que este gran objetivo plantea, hemos dedicado especial atención al problema del conocimiento autónomo del mundo. Nuestra mayor preocupación ha sido la de establecer mecanismos que permitan, a partir de información sensorial cruda, el desarrollo incremental de un modelo topológico del entorno en el que se mueve el robot. Estos mecanismos se apoyan invariablemente en un nuevo concepto propuesto en esta tesis: el gradiente sensorial. El gradiente sensorial es un dispositivo matemático que funciona como un detector de sucesos interesantes para el sistema. Una vez detectado uno de estos sucesos, el robot puede identificar su situación en un mapa topológico y actuar en consecuencia. Hemos denominado a estas situaciones especiales lugares sensorialmente relevantes, ya que (a) captan la atención del sistema y (b) pueden ser identificadas utilizando la información sensorial. Para explotar convenientemente los modelos construidos, hemos desarrollado un algoritmo capaz de elaborar planes internalizados, estableciendo una red de sugerencias en los lugares sensorialmente relevantes, de modo que el robot encuentra en estos puntos una dirección recomendada de navegación. Finalmente, hemos implementado un sistema de navegación robusto con habilidades para interpretar y adecuar los planes internalizados a las circunstancias concretas del momento. Nuestro sistema de navegación está basado en la teoría de campos de potencial artificial, a la que hemos incorporado la posibilidad de añadir cargas ficticias como ayuda a la evitación de mínimos locales. Como aportación adicional de esta tesis al campo genérico de la ciencia cognitiva, todos estos elementos se integran en una arquitectura centrada en la memoria, lo que pretende resaltar la importancia de ésta en los procesos cognitivos de los seres vivos y aporta un giro conceptual al punto de vista tradicional, centrado en los procesos. The general objective of this thesis is the development of a global navigation system endowed with planning and learning features for a mobile robot. Within this general objective we have devoted a special effort to the autonomous learning problem. Our main concern has been to establish the necessary mechanisms for the incremental development of a topological model of the robot’s environment using the sensory information. These mechanisms are based on a new concept proposed in the thesis: the sensory gradient. The sensory gradient is a mathematical device which works like a detector of “interesting” environment’s events. Once a particular event has been detected the robot can identify its situation in the topological map and to react accordingly. We have called these special situations relevant sensory places because (a) they capture the system’s attention and (b) they can be identified using the sensory information. To conveniently exploit the built-in models we have developed an algorithm able to make internalized plans, establishing a suggestion network in the sensory relevant places in such way that the robot can find at those places a recommended navigation direction. It has been also developed a robust navigation system able to navigate by means of interpreting and adapting the internalized plans to the concrete circumstances at each instant, i.e. a reactive navigation system. This reactive system is based on the artificial potential field approach with the additional feature introduced in the thesis of what we call fictitious charges as an aid to avoid local minima. As a general contribution of the thesis to the cognitive science field all the above described elements are integrated in a memory-based architecture, emphasizing the important role played by the memory in the cognitive processes of living beings and giving a conceptual turn in the usual process-based approach.

In-loop Feature Tracking for Structure and Motion with Out-of-core Optimization

In this paper, a novel and approach for obtaining 3D models from video sequences captured with hand-held cameras is addressed. We define a pipeline that robustly deals with different types of sequences and acquiring devices. Our system follows a divide and conquer approach: after a frame decimation that pre-conditions the input sequence, the video is split into short-length clips. This allows to parallelize the reconstruction step which translates into a reduction in the amount of computational resources required. The short length of the clips allows an intensive search for the best solution at each step of reconstruction which robustifies the system. The process of feature tracking is embedded within the reconstruction loop for each clip as opposed to other approaches. A final registration step, merges all the processed clips to the same coordinate frame

Técnicas de análisis, caracterización y detección de señales de voz en entornos acústicos adversos

Este trabajo de Tesis ha abordado el objetivo de dar robustez y mejorar la Detección de Actividad de Voz en entornos acústicos adversos con el fin de favorecer el comportamiento de muchas aplicaciones vocales, por ejemplo aplicaciones de telefonía basadas en reconocimiento automático de voz, aplicaciones en sistemas de transcripción automática, aplicaciones en sistemas multicanal, etc. En especial, aunque se han tenido en cuenta todos los tipos de ruido, se muestra especial interés en el estudio de las voces de fondo, principal fuente de error de la mayoría de los Detectores de Actividad en la actualidad. Las tareas llevadas a cabo poseen como punto de partida un Detector de Actividad basado en Modelos Ocultos de Markov, cuyo vector de características contiene dos componentes: la energía normalizada y la variación de la energía. Las aportaciones fundamentales de esta Tesis son las siguientes: 1) ampliación del vector de características de partida dotándole así de información espectral, 2) ajuste de los Modelos Ocultos de Markov al entorno y estudio de diferentes topologías y, finalmente, 3) estudio e inclusión de nuevas características, distintas de las del punto 1, para filtrar los pulsos de pronunciaciones que proceden de las voces de fondo. Los resultados de detección, teniendo en cuenta los tres puntos anteriores, muestran con creces los avances realizados y son significativamente mejores que los resultados obtenidos, bajo las mismas condiciones, con otros detectores de actividad de referencia. This work has been focused on improving the robustness at Voice Activity Detection in adverse acoustic environments in order to enhance the behavior of many vocal applications, for example telephony applications based on automatic speech recognition, automatic transcription applications, multichannel systems applications, and so on. In particular, though all types of noise have taken into account, this research has special interest in the study of pronunciations coming from far-field speakers, the main error source of most activity detectors today. The tasks carried out have, as starting point, a Hidden Markov Models Voice Activity Detector which a feature vector containing two components: normalized energy and delta energy. The key points of this Thesis are the following: 1) feature vector extension providing spectral information, 2) Hidden Markov Models adjustment to environment and study of different Hidden Markov Model topologies and, finally, 3) study and inclusion of new features, different from point 1, to reject the pronunciations coming from far-field speakers. Detection results, taking into account the above three points, show the advantages of using this method and are significantly better than the results obtained under the same conditions by other well-known voice activity detectors.

Two Different Approaches of Feature Extraction for Classifying the EEG Signals

The electroencephalograph (EEG) signal is one of the most widely used signals in the biomedicine field due to its rich information about human tasks. This research study describes a new approach based on i) build reference models from a set of time series, based on the analysis of the events that they contain, is suitable for domains where the relevant information is concentrated in specific regions of the time series, known as events. In order to deal with events, each event is characterized by a set of attributes. ii) Discrete wavelet transform to the EEG data in order to extract temporal information in the form of changes in the frequency domain over time- that is they are able to extract non-stationary signals embedded in the noisy background of the human brain. The performance of the model was evaluated in terms of training performance and classification accuracies and the results confirmed that the proposed scheme has potential in classifying the EEG signals.

Two Different Approaches of Feature Extraction for Classifying the EEG Signals

The focus of this chapter is to study feature extraction and pattern classification methods from two medical areas, Stabilometry and Electroencephalography (EEG). Stabilometry is the branch of medicine responsible for examining balance in human beings. Balance and dizziness disorders are probably two of the most common illnesses that physicians have to deal with. In Stabilometry, the key nuggets of information in a time series signal are concentrated within definite time periods are known as events. In this chapter, two feature extraction schemes have been developed to identify and characterise the events in Stabilometry and EEG signals. Based on these extracted features, an Adaptive Fuzzy Inference Neural network has been applied for classification of Stabilometry and EEG signals.

Evaluation of APD and SiPM Matrices as Sensors for Monolithic PET Detector Block

Gamma detectors based on monolithic scintillator blocks coupled to APDs matrices have proved to be a good alternative to pixelated ones for PET scanners. They provide comparable spatial resolution, improve the sensitivity and make easier the mechanical design of the system. In this study we evaluate by means of Geant4-based simulations the possibility of replacing the APDs by SiPMs. Several commercial matrices of light sensors coupled to LYSO:Ce monolithic blocks have been simulated and compared. Regarding the spatial resolution and linearity of the detector, SiPMs with high photo detection efficiency could become an advantageous replacement for the APDs

Evaluation of a PET Prototype Using LYSO:Ce Monolithic Detector Block

We have analyzed the performance of a PET demonstrator formed by two sectors of four monolithic detector blocks placed face-to-face. Both front-end and read-out electronics have been evaluated by means of coincidence measurements using a rotating 22Na source placed at the center of the sectors in order to emulate the behavior of a complete full ring. A continuous training method based on neural network (NN) algorithms has been carried out to determine the entrance points over the surface of the detectors. Reconstructed images from 1 MBq 22Na point source and 22Na Derenzo phantom have been obtained using both filtered back projection (FBP) analytic methods and the OSEM 3D iterative algorithm available in the STIR software package [1]. Preliminary data on image reconstruction from a 22Na point source with Ø = 0.25 mm show spatial resolutions from 1.7 to 2.1 mm FWHM in the transverse plane. The results confirm the viability of this design for the development of a full-ring brain PET scanner compatible with magnetic resonance imaging for human studies.

Signal analysis and feature generation for pattern identification of partial discharges in high-voltage equipment

This paper proposes a method for the identification of different partial discharges (PDs) sources through the analysis of a collection of PD signals acquired with a PD measurement system. This method, robust and sensitive enough to cope with noisy data and external interferences, combines the characterization of each signal from the collection, with a clustering procedure, the CLARA algorithm. Several features are proposed for the characterization of the signals, being the wavelet variances, the frequency estimated with the Prony method, and the energy, the most relevant for the performance of the clustering procedure. The result of the unsupervised classification is a set of clusters each containing those signals which are more similar to each other than to those in other clusters. The analysis of the classification results permits both the identification of different PD sources and the discrimination between original PD signals, reflections, noise and external interferences. The methods and graphical tools detailed in this paper have been coded and published as a contributed package of the R environment under a GNU/GPL license.

An Interval-based Multiobjective Approach to Feature Subset Selection Using Joint Modeling of Objectives and Variables

This paper studies feature subset selection in classification using a multiobjective estimation of distribution algorithm. We consider six functions, namely area under ROC curve, sensitivity, specificity, precision, F1 measure and Brier score, for evaluation of feature subsets and as the objectives of the problem. One of the characteristics of these objective functions is the existence of noise in their values that should be appropriately handled during optimization. Our proposed algorithm consists of two major techniques which are specially designed for the feature subset selection problem. The first one is a solution ranking method based on interval values to handle the noise in the objectives of this problem. The second one is a model estimation method for learning a joint probabilistic model of objectives and variables which is used to generate new solutions and advance through the search space. To simplify model estimation, l1 regularized regression is used to select a subset of problem variables before model learning. The proposed algorithm is compared with a well-known ranking method for interval-valued objectives and a standard multiobjective genetic algorithm. Particularly, the effects of the two new techniques are experimentally investigated. The experimental results show that the proposed algorithm is able to obtain comparable or better performance on the tested datasets.

Neutron spectrometry using artificial neural networks for a bonner sphere spectrometer with 3He detector

Neutron spectra unfolding and dose equivalent calculation are complicated tasks in radiation protection, are highly dependent of the neutron energy, and a precise knowledge on neutron spectrometry is essential for all dosimetry-related studies as well as many nuclear physics experiments. In previous works have been reported neutron spectrometry and dosimetry results, by using the ANN technology as alternative solution, starting from the count rates of a Bonner spheres system with a LiI(Eu) thermal neutrons detector, 7 polyethylene spheres and the UTA4 response matrix with 31 energy bins. In this work, an ANN was designed and optimized by using the RDANN methodology for the Bonner spheres system used at CIEMAT Spain, which is composed of a He neutron detector, 12 moderator spheres and a response matrix for 72 energy bins. For the ANN design process a neutrons spectra catalogue compiled by the IAEA was used. From this compilation, the neutrons spectra were converted from lethargy to energy spectra. Then, the resulting energy ?uence spectra were re-binned by using the MCNP code to the corresponding energy bins of the He response matrix before mentioned. With the response matrix and the re-binned spectra the counts rate of the Bonner spheres system were calculated and the resulting re-binned neutrons spectra and calculated counts rate were used as the ANN training data set.

Time-Frequency based Feature Selection for Discrimination of non stationary Biosignals.

This research proposes a generic methodology for dimensionality reduction upon time-frequency representations applied to the classification of different types of biosignals. The methodology directly deals with the highly redundant and irrelevant data contained in these representations, combining a first stage of irrelevant data removal by variable selection, with a second stage of redundancy reduction using methods based on linear transformations. The study addresses two techniques that provided a similar performance: the first one is based on the selection of a set of the most relevant time?frequency points, whereas the second one selects the most relevant frequency bands. The first methodology needs a lower quantity of components, leading to a lower feature space; but the second improves the capture of the time-varying dynamics of the signal, and therefore provides a more stable performance. In order to evaluate the generalization capabilities of the methodology proposed it has been applied to two types of biosignals with different kinds of non-stationary behaviors: electroencephalographic and phonocardiographic biosignals. Even when these two databases contain samples with different degrees of complexity and a wide variety of characterizing patterns, the results demonstrate a good accuracy for the detection of pathologies, over 98%.The results open the possibility to extrapolate the methodology to the study of other biosignals.

A new feature extraction method for signal classification applied to cat spinal cord signals

In the spinal cord of the anesthetized cat, spontaneous cord dorsum potentials (CDPs) appear synchronously along the lumbo-sacral segments. These CDPs have different shapes and magnitudes. Previous work has indicated that some CDPs appear to be specially associated with the activation of spinal pathways that lead to primary afferent depolarization and presynaptic inhibition. Visual detection and classification of these CDPs provides relevant information on the functional organization of the neural networks involved in the control of sensory information and allows the characterization of the changes produced by acute nerve and spinal lesions. We now present a novel feature extraction approach for signal classification, applied to CDP detection. The method is based on an intuitive procedure. We first remove by convolution the noise from the CDPs recorded in each given spinal segment. Then, we assign a coefficient for each main local maximum of the signal using its amplitude and distance to the most important maximum of the signal. These coefficients will be the input for the subsequent classification algorithm. In particular, we employ gradient boosting classification trees. This combination of approaches allows a faster and more accurate discrimination of CDPs than is obtained by other methods.

Feature selection in the reconstruction of complex network representations of spectral data

Complex networks have been extensively used in the last decade to characterize and analyze complex systems, and they have been recently proposed as a novel instrument for the analysis of spectra extracted from biological samples. Yet, the high number of measurements composing spectra, and the consequent high computational cost, make a direct network analysis unfeasible. We here present a comparative analysis of three customary feature selection algorithms, including the binning of spectral data and the use of information theory metrics. Such algorithms are compared by assessing the score obtained in a classification task, where healthy subjects and people suffering from different types of cancers should be discriminated. Results indicate that a feature selection strategy based on Mutual Information outperforms the more classical data binning, while allowing a reduction of the dimensionality of the data set in two orders of magnitude

Mining recurring concepts in a dynamic feature space

Most data stream classification techniques assume that the underlying feature space is static. However, in real-world applications the set of features and their relevance to the target concept may change over time. In addition, when the underlying concepts reappear, reusing previously learnt models can enhance the learning process in terms of accuracy and processing time at the expense of manageable memory consumption. In this paper, we propose mining recurring concepts in a dynamic feature space (MReC-DFS), a data stream classification system to address the challenges of learning recurring concepts in a dynamic feature space while simultaneously reducing the memory cost associated with storing past models. MReC-DFS is able to detect and adapt to concept changes using the performance of the learning process and contextual information. To handle recurring concepts, stored models are combined in a dynamically weighted ensemble. Incremental feature selection is performed to reduce the combined feature space. This contribution allows MReC-DFS to store only the features most relevant to the learnt concepts, which in turn increases the memory efficiency of the technique. In addition, an incremental feature selection method is proposed that dynamically determines the threshold between relevant and irrelevant features. Experimental results demonstrating the high accuracy of MReC-DFS compared with state-of-the-art techniques on a variety of real datasets are presented. The results also show the superior memory efficiency of MReC-DFS.

Implementing the weakest failure detector for solving consensus

The concept of unreliable failure detector was introduced by Chandra and Toueg as a mechanism that provides information about process failures. This mechanism has been used to solve several agreement problems, such as the consensus problem. In this paper, algorithms that implement failure detectors in partially synchronous systems are presented. First two simple algorithms of the weakest class to solve the consensus problem, namely the Eventually Strong class (⋄S), are presented. While the first algorithm is wait-free, the second algorithm is f-resilient, where f is a known upper bound on the number of faulty processes. Both algorithms guarantee that, eventually, all the correct processes agree permanently on a common correct process, i.e. they also implement a failure detector of the class Omega (Ω). They are also shown to be optimal in terms of the number of communication links used forever. Additionally, a wait-free algorithm that implements a failure detector of the Eventually Perfect class (⋄P) is presented. This algorithm is shown to be optimal in terms of the number of bidirectional links used forever.